Sie sind auf Seite 1von 9

Big Data: How to Handle: A survey

Dinesh ABC
MCA Deptt. MCA Deptt.
PDM University, Bahadurgarh

Abstract— industries [4], which are broadly built on the third platform,
Todays Information environment produces a large amount of data, that mainly referring to big data, cloud computing, internet of
need to be Handle, so that a Knowledge can be get from things, and social business. Generally, Data warehouses have
In this new era with boom of data both structured and unstructured,
in the field of genomics, meteorology, biology, environmental research been used to manage the large dataset. In this case extracting
and many others, the precise knowledge from the available big data is a foremost
it has become difficult to process, manage and analyze Data issue. Most of the presented approaches in data mining are not
using traditional Techniques. So, some Proper tools to be used to usually able to handle the large datasets successfully. The key
handleand should be understood to gain knowledge about the problem in the analysis of big data is the lack of coordination
Big Data. This paper presents a review of various algorithms between database systems as well as with analysis tools such as
necessary for handling Big Data. Explore different stages and solution data mining and statistical analysis. These challenges generally
given to handle it. arise when we wish to perform knowledge discovery and repre-
. sentation for its practical applications. A fundamental problem
is how to quantitatively describe the essential characteristics
Keywords: Big Data, Apache, Hadoop
of big data. There is a need for epistemological implications
in describing data revolution [5]. Additionally, the study on
complexity theory of big data will help understand essential
I. INTRODUCTION characteristics and formation of complex patterns in big data,
simplify its representation, gets better knowledge abstraction,
In digital world, data are generated from various sources and guide the design of computing models and algorithms
and the fast transition from digital technologies has led to on big data [4]. Much research was carried out by various
growth of big data. It provides evolutionary breakthroughs in researchers on big data and its trends [6], [7], [8].
many fields with collection of large datasets. In general, it
refers to the collection of large and complex datasets which However, it is to be noted that all data available in the
are difficult to process using traditional database management form of big data are not useful for analysis or decision making
tools or data processing applications. These are available process. Industry and academia are interested in disseminating
in structured, semi-structured, and unstructured format in the findings of big data. This paper focuses on challenges in
petabytes and beyond. Formally, it is defined from 3Vs to 4Vs. big data and its available techniques. Additionally, we state
3Vs refers to volume, velocity, and variety. Volume refers to open research issues in big data. So, to elaborate this, the
the huge amount of data that are being generated everyday paper is divided into following sections. Sections 2 deals
whereas velocity is the rate of growth and how fast the data with challenges that arise during fine tuning of big data.
are gathered for being analysis. Variety provides information Section 3 furnishes the open research issues that will help
about the types of data such as structured, unstructured, semi- us to process big data and extract useful knowledge from it.
structured etc. The fourth V refers to veracity that includes Section 4 provides an insight to big data tools and techniques.
availability and accountability. The prime objective of big data Conclusion remarks are provided in section 5 to summarize
analysis is to process data of high volume, velocity, variety, and outcomes.
veracity using various traditional and computational intelligent
techniques [1]. Some of these extraction methods for obtaining II. CHALLENGES IN Large Data Set ANALYTICS
helpful information was discussed by Gandomi and Haider
Recent years big data has been accumulated in several
[2]. The following Figure 1 refers to the definition of big
data. However exact definition for big data is not defined and domains like health care, public administration, retail, bio-
there is a believe that it is problem specific. This will help us chemistry, and other interdisciplinary scientific researches.
in obtaining enhanced decision making, insight discovery and Web-based applications encounter big data frequently, such
optimization while being innovative and cost-effective. as social computing, internet text and documents, and inter-
net search indexing. Social computing includes social net-
It is expected that the growth of big data is estimated to work analysis, online communities, recommender systems,
reach 25 billion by 2015 [3]. From the perspective of the reputation systems, and prediction markets where as internet
information and communication technology, big data is a ro- search indexing includes ISI, IEEE Xplorer, Scopus, Thomson
bust impetus to the next generation of information technology
 

unprecedented challenge for researchers. It is becuase, existing


algorithms may not always respond in an adequate time when
dealing with these high dimensional data. Automation of this
process and developing new machine learning algorithms to
ensure consistency is a major challenge in recent years. In
addition to all these Clustering of large datasets that help
Fig. 1: Characteristics of Big Data in analyzing the big data is of prime concern [11]. Recent
technologies such as hadoop and mapReduce make it possible
to collect large amount of semi structured and unstructured
data in a reasonable amount of time. The key engineering
challenge is how to effectively analyze these data for obtaining
better knowledge. A standard process to this end is to transform
the semi structured or unstructured data into structured data,
and then apply data mining algorithms to extract knowledge. A
framework to analyze data was discussed by Das and Kumar
[12]. Similarly detail explanation of data analysis for public
tweets was also discussed by Das et al in their paper .
The major challenge in this case is to pay more attention for
designing storage sytems and to elevate efficient data analysis
Reuters etc. Considering this advantages of big data it provides tool that provide guarantees on the output when the data
a new opportunities in the knowledge processing tasks for comes from different sources. Furthermore, design of machine
the upcoming researchers. However oppotunities always follow
learning algorithms to analyze data is essential for improving
some challenges.
efficiency and scalability.
To handle the challenges we need to know various compu-
tational complexities, information security, and computational B. Knowledge Discovery and Computational Complexities
method, to analyze big data. For example, many statistical Knowledge discovery and representation is a prime issue
methods that perform well for small data size do not scale in big data. It includes a number of sub fields such as
to voluminous data. Similarly, many computational techniques authentication, archiving, management, preservation, informa-
that perform well for small data face significant challenges in tion retrieval, and representation. There are several tools for
analyzing big data. Various challenges that the health sector knowledge discovery and representation such as fuzzy set
face was being researched by much researchers [9], [10]. Here , rough set , soft set , near set , formal
the challenges of big data analytics are classified into four concept analysis [18], principal component analysis [19] etc to
broad categories namely data storage and analysis; knowledge name a few. Additionally many hybridized techniques are also
discovery and computational complexities; scalability and vi- developed to process real life problems. All these techniques
sualization of data; and information security. We discuss these are problem dependent. Further some of these techniques may
issues briefly in the following subsections. not be suitable for large datasets in a sequential computer. At
the same time some of the techniques has good characteristics
A. Data Storage and Analysis of scalability over parallel computer. Since the size of big
In recent years the size of data has grown exponentially data keeps increasing exponentially, the available tools may
by various means such as mobile devices, aerial sensory not be efficient to process these data for obtaining meaningful
technologies, remote sensing, radio frequency identification information. The most popular approach in case of larage
readers etc. These data are stored on spending much cost dataset management is data warehouses and data marts. Data
whereas they ignored or deleted finally becuase there is no warehouse is mainly responsible to store data that are sourced
enough space to store them. Therefore, the first challenge for from operational systems whereas data mart is based on a data
big data analysis is storage mediums and higher input/output warehouse and facilitates analysis.
speed. In such cases, the data accessibility must be on the Analysis of large dataset requires more computational
top priority for the knowledge discovery and representation. complexities. The major issue is to handle inconsistencies
The prime reason is being that, it must be accessed easily and and uncertainty present in the datasets. In general, systematic
promptly for further analysis. In past decades, analyst use hard modeling of the computational complexity is used. It may be
disk drives to store data but, it slower random input/output difficult to establish a comprehensive mathematical system that
performance than sequential input/output. To overcome this is broadly applicable to Big Data. But a domain specific data
limitation, the concept of solid state drive (SSD) and phrase analytics can be done easily by understanding the particular
change memory (PCM) was introduced. However the avialable complexities. A series of such development could simulate big
storage technologies cannot possess the required performance data analytics for different areas. Much research and survey
for processing big data. has been carried out in this direction using machine learning
Another challenge with Big Data analysis is attributed techniques with the least memory requirements. The basic
objective in these research is to minimize computational cost
to diversity of data. with the ever growing of datasets, data
processing and complexities .
mining tasks has significantly increased. Additionally data
reduction, data selection, feature selection is an essential task However, current big data analysis tools have poor per-
especially when dealing with large datasets. This presents an formance in handling computational complexities, uncertainty,
and inconsistencies. It leads to a great challenge to develop III. OPEN RESEARCH ISSUES IN BIG DATA ANALYTICS
techniques and technologies that can deal computational com-
Big data analytics and data science are becoming the
plexity, uncertainty,and inconsistencies in a effective manner.
research focal point in industries and academia. Data science
aims at researching big data and knowledge extraction from
C. Scalability and Visualization of Data data. Applications of big data and data science include infor-
The most important challenge for big data analysis tech- mation science, uncertainty modeling, uncertain data analysis,
machine learning, statistical learning, pattern recognition, data
niques is its scalability and security. In the last decades
warehousing, and signal processing. Effective integration of
researchers have paid attentions to accelerate data analysis and
technologies and analysis will result in predicting the future
its speed up processors followed by Moore’s Law. For the
drift of events. Main focus of this section is to discuss open
former, it is necessary to develop sampling, on-line, and mul-
research issues in big data analytics. The research issues
tiresolution analysis techniques. Incremental techniques have
pertaining to big data analysis are classified into three broad
good scalability property in the aspect of big data analysis. As
categories namely internet of things (IoT), cloud computing,
the data size is scaling much faster than CPU speeds, there is a
bio inspired computing, and quantum computing. However it
natural dramatic shift in processor technology being embedded
is not limited to these issues. More research issues related to
with increasing number of cores . This shift in processors
health care big data can be found in Husing Kuo et al. paper
leads to the development of parallel computing. Real time
[9].
applications like navigation, social networks, finance, internet
search, timeliness etc. requires parallel computing.
A. IoT for Big Data Analytics
The objective of visualizing data is to present them more
adequately using some techniques of graph theory. Graphical Internet has restructured global interrelations, the art of
visualization provides the link between data with proper inter- businesses, cultural revolutions and an unbelievable number
pretation. However, online marketplace like flipkart, amazon, of personal characteristics. Currently, machines are getting in
e-bay have millions of users and billions of goods to sold each on the act to control innumerable autonomous gadgets via
month. This generates a lot of data. To this end, some company internet and create Internet of Things (IoT). Thus, appliances
uses a tool Tableau for big data visualization. It has capability are becoming the user of the internet, just like humans with
to transform large and complex data into intuitive pictures. This the web browsers. Internet of Things is attracting the attention
help employees of a company to visualize search relevance, of recent researchers for its most promising opportunities
monitor latest customer feeback, and their sentiment analysis. and challenges. It has an imperative economic and societal
However, current big data visualization tools mostly have poor impact for the future construction of information, network and
performances in functionalities, scalability, and response in communication technology. The new regulation of future will
time. be eventually, everything will be connected and intelligently
controlled. The concept of IoT is becoming more pertinent
We can observe that big data have produced many chal- to the realistic world due to the development of mobile de-
lenges for the developments of the hardware and software vices, embedded and ubiquitous communication technologies,
which leads to parallel computing, cloud computing, dis- cloud computing, and data analytics. Moreover, IoT presents
tributed computing, visualization process, scalability. To over- challenges in combinations of volume, velocity and variety.
come this issue, we need to correlate more mathematical In a broader sense, just like the internet, Internet of Things
models to computer science. enables the devices to exist in a myriad of places and facilitates
applications ranging from trivial to the crucial. Conversely, it is
D. Information Security still mystifying to understand IoT well, including definitions,
content and differences from other similar concepts. Several
In big data analysis massive amount of data are correlated, diversified technologies such as computational intelligence,
analyzed, and mined for meaningful patterns. All organizations and big-data can be incorporated together to improve the
have different policies to safe guard their sensitive information. data management and knowledge discovery of large scale
Preserving sensitive information is a major issue in big data automation applications. Much research in this direction has
analysis. There is a huge security risk associated with big data been carried out by Mishra, Lin and Chang .
. Therefore, information security is becoming a big data
Knowledge acquisition from IoT data is the biggest chal-
analytics problem. Security of big data can be enhanced by
using the techniques of authentication, authorization, and en- lenge that big data professional are facing. Therefore, it is
cryption. Various security measures that big data applications essential to develop infrastructure to analyze the IoT data. An
face are scale of network, variety of different devices, real time IoT device generates continuous streams of data and the re-
security monitoring, and lack of intrusion system . searchers can develop tools to extract meaningful information
The security challenge caused by big data has attracted the from these data using machine learning techniques. Under-
attention of information security. Therefore, attention has to standing these streams of data generated from IoT devices and
be given to develop a multi level security policy model and analysing them to get meaningful information is a challenging
prevention system. issue and it leads to big data analytics. Machine learning
algorithms and computational intelligence techniques is the
Although much research has been carried out to secure only solution to handle big data from IoT prospective. Key
big data [25] but it requires lot of improvement. The major technologies that are associated with IoT are also discussed in
challenge is to develop a multi-level security, privacy preserved many research papers [28]. Figure 2 depicts an overview of
data model for big data. IoT big data and knowledge discovery process.
 

infrastructures that are hidden in virtualization software make


systems to behave like a true computer, but with the flexibility
of specification details such as number of processors, disk
space, memory, and operating system. The use of these virtual
computers is known as cloud computing which has been one
Fig. 2: IoT Big Data Knowledge Discovery of the most robust big data technique. Big Data and cloud
computing technologies are developed with the importance of
developing a scalable and on demand availability of resources
and data. Cloud computing harmonize massive data by on-
demand access to configurable computing resources through
virtualization techniques. The benefits of utilizing the Cloud
computing include offering resources when there is a demand
and pay only for the resources which is needed to develop
the product. Simultaneously, it improves availability and cost
reduction. Open challenges and research issues of big data
and cloud computing are discussed in detail by many re-
Knowledge exploration system have originated from theo- searchers which highlights the challenges in data management,
ries of human information processing such as frames, rules, data variety and velocity, data storage, data processing, and
tagging, and semantic networks. In general, it consists of resource management . So Cloud computing helps
four segments such as knowledge acquisition, knowledge in developing a business model for all varieties of applications
base, knowledge dissemination, and knowledge application. with infrastructure and tools.
In knowledge acquisition phase, knowledge is discovered Big data application using cloud computing should support
by using various traditional and computational intelligence data analytic and development. The cloud environment should
techniques. The discovered knowledge is stored in knowledge provide tools that allow data scientists and business analysts to
bases and expert systems are generally designed based on the interactively and collaboratively explore knowledge acquisition
discovered knowledge. Knowledge dissemination is important data for further processing and extracting fruitful results.
for obtaining meaningful information from the knowledge This can help to solve large applications that may arise in
base. Knowledge extraction is a process that searches doc- various domains. In addition to this, cloud computing should
uments, knowledge within documents as well as knowledge also enable scaling of tools from virtual technologies into
bases. The final phase is to apply discovered knowledge in new technologies like spark, R, and other types of big data
various applications. It is the ultimate goal of knowledge processing techniques.
discovery. The knowledge exploration system is necessarily
iterative with the judgement of knowledge application. There Big data forms a framework for discussing cloud comput-
are many issues, discussions, and researches in this area of ing options. Depending on special need, user can go to the
knowledge exploration. It is beyond scope of this survey marketplace and buy infrastructure services from cloud service
paper. For better visualization, knowledge exploration system providers such as Google, Amazon, IBM, software as a service
is depicted in Figure 3. (SaaS) from a whole crew of companies such as NetSuite,
Cloud9, Jobscience etc. Another advantage of cloud computing
is cloud storage which provides a possible way for storing big
data. The obvious one is the time and cost that are needed to
upload and download big data in the cloud environment. Else,
it becomes difficult to control the distribution of computation
and the underlying hardware. But, the major issues are privacy
concerns relating to the hosting of data on public servers,
and the storage of data from human studies. All these issues
will take big data and cloud computing to a high level of
development.

C. Bio-inspired Computing for Big Data Analytics


Bio-inspired computing is a technique inspired ny nature
to address complex real world problems. Biological systems
are self organized without a central control. A bio-inspired
cost minimization mechanism search and find the optimal
Fig. 3: IoT Knowledge Exploration System data service solution on considering cost of data management
and service maintenance. These techniques are developed by
biological molecules such as DNA and proteins to conduct
computational calculations involving storing, retrieving, and
B. Cloud Computing for Big Data Analytics processing of data. A significant feature of such computing
is that it integrates biologically derived materials to perform
The development of virtualization technologies have made computational functions and receive intelligent performance.
supercomputing more accessible and affordable. Computing These systems are more suitable for big data applications.
Huge amount of data are generated from variety of resources
across the web since the digitization. Analyzing these data
and categorizing into text, image and video etc will require For example Dremel and Apache Drill are the big data plat-
lot of intelligent analytics from data scientists and big data forms that support interactive analysis. These tools help us in
professionals. Proliferations of technologies are emerging like developing the big data projects. A fabulous list of big data
big data, IoT, cloud computing, bio inspired computing etc tools and techniques is also discussed by much researchers [6],
whereas equilibrium of data can be done only by selecting right [34]. The typical work flow of big data project discussed by
platform to analyze large and furnish cost effective results. Huang et al is highlighted in this section [35] and is depicted
in Figure 4.
Bio-inspired computing techniques serve as a key role in
intelligent data analysis and its application to big data. These
algorithms help in performing data mining for large datasets
due to its optimization application. The most advantage is its
simplicity and their rapid concergence to optimal solution
while solving service provision problems. Some applications to
this end using bio inspired computing was discussed in detail
by Cheng et al [32]. From the discussions, we can observe that
the bio-inspired computing models provide smarter interac-
tions, inevitable data losses, and help is handling ambiguities.
Hence, it is believed that in future bio-inspired computing may
help in handling big data to a large extent.

D. Quantum Computing for Big Data Analysis


A quantum computer has memory that is exponentially Fig. 4: Workflow of Big Data Project
larger than its physical size and can manipulate an exponential
set of inputs simultaneously [33]. This exponential improve-
ment in computer systems might be possible. If a real quantum
computer is available now, it could have solved problems A. Apache Hadoop and MapReduce
that are exceptionally difficult on recent computers, of course The most established software platform for big data anal-
today’s big data problems. The main technical difficulty in ysis is Apache Hadoop and Mapreduce. It consists of hadoop
building quantum computer could soon be possible. Quantum kernel, mapreduce, hadoop distributed file system (HDFS)
computing provides a way to merge the quantum mechanics to and apache hive etc. Map reduce is a programming model
process the information. In traditional computer, information for processing large datasets is based on divide and conquer
is presented by long strings of bits which encode either a method. The divide and conquer method is implemented in two
zero or a one. On the other hand a quantum computer uses steps such as Map step and Reduce Step. Hadoop works on
quantum bits or qubits. The difference between qubit and bit two kinds of nodes such as master node and worker node. The
is that, a qubit is a quantum system that encodes the zero and master node divides the input into smaller sub problems and
the one into two distinguishable quantum states. Therefore, then distributes them to worker nodes in map step. Thereafter
it can be capitalized on the phenomena of superposition and the master node combines the outputs for all the subproblems
entanglement. It is because qubits behave quantumly. For in reduce step. Moreover, Hadoop and MapReduce works as
example, 100 qubits in quantum systems require 2100 complex a powerful software framework for solving big data problems.
values to be stored in a classic computer system. It means that It is also helpful in fault-tolerant storage and high throughput
many big data problems can be solved much faster by larger data processing.
scale quantum computers compared with classical computers.
Hence it is a challenge for this generation to built a quantum B. Apache Mahout
computer and facilitate quantum computing to solve big data
Apache mahout aims to provide scalable and commercial
problems.
machine learning techniques for large scale and intelligent data
analysis applications. Core algorithms of mahout including
IV. TOOLS FOR BIG DATA PROCESSING clustering, classification, pattern mining, regression, dimen-
Large numbers of tools are available to process big data. In
sionalty reduction, evolutionary algorithms, and batch based
collaborative filtering run on top of Hadoop platform through
this section, we discuss some current techniques for analyzing map reduce framework. The goal of mahout is to build a
big data with emphasis on three important emerging tools vibrant, responsive, diverse community to facilitate discussions
namely MapReduce, Apache Spark, and Storm. Most of the on the project and potential use cases. The basic objective
available tools concentrate on batch processing, stream pro- of Apache mahout is to provide a tool for elleviating big
cessing,and interactive analysis. Most batch processing tools challenges. The different companies those who have imple-
are based on the Apache Hadoop infrastructure such as Mahout mented scalable machine learning algorithms are Google, IBM,
and Dryad. Stream data applications are mostly used for real Amazon, Yahoo, Twitter, and facebook .
time analytic. Some examples of large scale streaming platform
are Strom and Splunk. The interactive analysis process allow
C. Apache Spark
users to directly interact in real time for their own analysis.
Apache spark is an open source big data processing frame-
work built for speed processing, and sophisticated analytics.
It is easy to use and was originally developed in 2009 in UC D. Dryad
Berkeleys AMPLab. It was open sourced in 2010 as an Apache
It is another popular programming model for implementing
project. Spark lets you quickly write applications in java, scala,
parallel and distributed programs for handling large context
or python. In addition to map reduce operations, it supports
bases on dataflow graph. It consists of a cluster of computing
SQL queries, streaming data, machine learning, and graph data
nodes, and an user use the resources of a computer cluster
processing. Spark runs on top of existing hadoop distributed
to run their program in a distributed way. Indeed, a dryad
file system (HDFS) infrastructure to provide enhanced and
user use thousands of machines, each of them with multiple
additional functionality. Spark consists of components namely
processors or cores. The major advantage is that users do
driver program, cluster manager and worker nodes. The driver
not need to know anything about concurrent programming. A
program serves as the starting point of execution of an appli-
dryad application runs a computational directed graph that is
cation on the spark cluster. The cluster manager allocates the
composed of computational vertices and communication chan-
resources and the worker nodes to do the data processing in
nels. Therefore, dryad provides a large number of functionality
the form of tasks. Each application will have a set of processes
including generating of job graph, scheduling of the machines
called executors that are responsible for executing the tasks.
for the available processes, transition failure handling in the
The major advantage is that it provides support for deploying
cluster, collection of performance metrics, visualizing the job,
spark applications in an existing hadoop clusters. Figure 5
invokinguser defined policies and dynamically updating the job
depicts the architecture diagram of Apache Spark. The various
graph in response to these policy decisions without knowing
features of Apache Spark are listed below:
the semantics of the vertices .

E. Storm
• The prime focus of spark includes resilient distributed
datasets (RDD), which store data in-memory and Storm is a distributed and fault tolerant real time com-
provide fault tolerance without replication. It supports putation system for processing large streaming data. It is
iterative computation, improves speed and resource specially designed for real time processing in contrasts with
utilization. hadoop which is for batch processing. Additionally, it is also
easy to set up and operate, scalable, fault-tolerant to provide
• The foremost advantage is that in addition to MapRe- competitive performances. The storm cluster is apparently
duce, it also supports streaming data, machine learn- similar to hadoop cluster. On storm cluster users run different
ing, and graph algorithms. topologies for different storm tasks whereas hadoop platform
implements map reduce jobs for corresponding applications.
• Another advantage is that, a user can run the applica-
There are number of differences between map reduce jobs
tion program in different languages such as Java, R, and topologies. The basic difference is that map reduce job
Python, or Scala. This is possible as it comes with eventually finishes whereas a topology processes messages all
higher-level libraries for advanced analytics. These the time, or until user terminate it. A storm cluster consists of
standard libraries increase developer productivity and two kinds of nodes such as master node and worker node. The
can be seamlessly combined to create complex work- master node and worker node implement two kinds of roles
flows. such as nimbus and supervisor respectively. The two roles have
• Spark helps to run an application in Hadoop cluster, similar functions in accordance with jobtracker and tasktracker
up to 100 times faster in memory, and 10 times faster of map reduce framework. Nimbus is in charge of distributing
when running on disk. It is possible because of the code across the storm cluster, scheduling and assigning tasks
reduction in number of read or write operations to
to worker nodes, and monitoring the whole system. The
disk. supervisor complies tasks as assigned to them by nimbus.
In addition, it start and terminate the process as necessary
• It is written in scala programming language and runs based on the instructions of nimbus. The whole computational
on java virtual machine (JVM) environment. Addi- technology is partitioned and distributed to a number of worker
tionally, it upports java, python and R for developing processes and each worker process implements a part of the
applications using Spark. topology.

F. Apache Drill
Apache drill is another distributed system for interactive
analysis of big data. It has more flexibility to support many
types of query languages, data formats, and data sources. It is
also specially designed to exploit nested data. Also it has an
objective to scale up on 10,000 servers or more and reaches the
capability to process patabytes of data and trillions of records
in seconds. Drill use HDFS for storage and map reduce to
perform batch analysis.

G. Jaspersoft
The Jaspersoft package is an open source software that
produce reports from database columns. It is a scalable big
data analytical platform and has a capability of fast data visu- issue. Expressing data access requirements of applications
alization on popular storage platforms, including MangoDB, and designing programming language abstractions to exploit
Cassandra, Redis etc. One important property of Jaspersoft parallelism are an immediate need .
is that it can quickly explore big data without extraction, Additionally, machine learning concepts and tools are
transformation, and loading (ETL). In addition to this, it also
gaining popularity among researchers to facilitate meaningful
have an ability to build powerful hypertext markup language
results from these concepts. Research in the area of machine
(HTML) reports and dashboards interactively and directly from
learning for big data has focused on data processing, algo-
big data store without ETL requirement. These generated
rithm implementation, and optimization. Many of the machine
reports can be shared with anyone inside or outside user’s
learning tools for big data are started recently needs drastic
organization.
change to adopt it. We argue that while each of the tools has
their advantages and limitations, more efficient tools can be
H. Splunk
developed for dealing with problems inherent to big data. The
In recent years a lot of data are generated through machine efficient tools to be developed must have provision to handle
from business industries. Splunk is a real-time and intelligent noisy and imbalance data, uncertainty and inconsistency, and
platform developed for exploiting machine generated big data. missing values.
It combines the up-to-the-moment cloud technologies and big
data. In turn it helps user to search, monitor, and analyze their VI. CONCLUSION
machine generated data through web interface. The results are In recent years a lot of data is generated .Analyzing these data is Not
exhibited in an intuitive way such as graphs, reports, and alerts.
Splunk is different from other stream processing tools. Its easy. In this paper, we survey the various research issues,
peculiarities include indexing structured, unstructured machine to analyze these big data. Here you can understand where to use batch
generated data, real-time searching, reporting analytical results, processing and where to use real time analytics. We specially use
and dashboards. The most important objective of Splunk is Hadoop, this is very important tool now a days to handle Big Data.
to provide metrices for many application, diagnose problems Different techniques used for the analysis include statistical analysis,
for system and information technology infrastructures, and machine learning, data mining, intelligent analysis, cloud com-
intelligent support for business operations. puting, quantum computing, and data stream processing.
In future researchers any body can pay more attention to
V. SUGGESTIONS FOR FUTURE WORK these techniques to solve problems of big data effectively and
The amount of data collected from various applications efficiently.
all over the world across a wide variety of fields today is
REFERENCES
expected to double every two years. It has no utility unless
these are analyzed to get useful information. This necessitates
the development of techniques which can be used to facilitate [1] Books Hadoop: The Definitive Guide 4th Edition
big data analysis. The development of powerful computers is Tom White and
a boon to implement these techniques leading to automated
systems. The transformation of data into knowledge is by Big Data: A Revolution That Will Transform How We Live, Work, and
no means an easy task for high performance large-scale data
Think by V. Mayer-Schönberger and K. Cukier
processing, including exploiting parallelism of current and
upcoming computer architectures for data mining. Moreover,
these data may involve uncertainty in many different forms.
Many different models like fuzzy sets, rough sets, soft sets, [2] A. Gandomi and M. Haider, Beyond the hype: Big data concepts, meth-
neural networks, their generalizations and hybrid models ob- ods, and analytics, International Journal of Information Management,
tained by combining two or more of these models have been 35(2) (2015), pp.137-144.
found to be fruitful in representing data. These models are [3] C. Lynch, Big data: How do your data grow?, Nature, 455 (2008),
also very much fruitful for analysis. More often than not, big pp.28-29.
data are reduced to include only the important characteristics [4] X. Jin, B. W.Wah, X. Cheng and Y. Wang, Significance and challenges
necessary from a particular study point of view or depending of big data research, Big Data Research, 2(2) (2015), pp.59-64.
upon the application area. So, reduction techniques have been [5] R. Kitchin, Big Data, new epistemologies and paradigm shifts, Big
Data Society, 1(1) (2014), pp.1-12.
developed. Often the data collected have missing values. These
[6] C. L. Philip, Q. Chen and C. Y. Zhang, Data-intensive applications,
values need to be generated or the tuples having these missing challenges, techniques and technologies: A survey on big data, Infor-
values are eliminated from the data set before analysis. More mation Sciences, 275 (2014), pp.314-347.
importantly, these new challenges may comprise, sometimes [7] K. Kambatla, G. Kollias, V. Kumar and A. Gram, Trends in big data
even deteriorate, the performance, efficiency and scalability analytics, Journal of Parallel and Distributed Computing, 74(7) (2014),
of the dedicated data intensive computing systems. The later pp.2561-2573.
approach sometimes leads to loss of information and hence [8] S. Del. Rio, V. Lopez, J. M. Bentez and F. Herrera, On the use of
not preferred. This brings up many research issues in the mapreduce for imbalanced big data using random forest, Information
Sciences, 285 (2014), pp.112-137.
industry and research community in forms of capturing and
[9] MH. Kuo, T. Sahama, A. W. Kushniruk, E. M. Borycki and D. K.
accessing data effectively. In addition, fast processing while
Grunwell, Health big data analytics: current perspectives, challenges
achieving high performance and high throughput, and storing and potential solutions, International Journal of Big Data Intelligence,
it efficiently for future use is another issue. Further, pro- 1 (2014), pp.114-126.
gramming for big data analysis is an important challenging [10]
.
[11] Z. Huang, A fast clustering algorithm to cluster very large categorical
data sets in data mining, SIGMOD Workshop on Research Issues on
Data Mining and Knowledge Discovery, 1997.